The association between Musculoskeletal Pain and Psychological Distress among five immigrant groups: Results of a cross-sectional study in Oslo, Norway

(1)

The association between Musculoskeletal Pain and Psychological Distress among five immigrant groups:

Results of a cross‐sectional study in Oslo, Norway

Steve Diaz French

Oslo 2009

This Thesis was written under the supervision of

Lars Lien MD, Ph.D.

Associate Professor Institute of General Practice and Community Medicine, Section for Preventive

Medicine and Epidemiology (FEPI) University of Oslo

Hein Stigum, Ph.D.

Senior Scientist Norwegian Institute of Public Health

A thesis submitted in partial fulfillment of the requirement for the degree of Master of Philosophy in International Community Health Institute of General Practice and Community Medicine, University of Oslo

(2)

2

Acknowledgements

First, I would like to thank my supervisor Lars Lien for his patience and support during this year. Lars´ motivation and dedication of his time, made it possible for me to meet the deadline. I am lucky to have you as my supervisor! Thank you very much!

Secondly, I would like to thank my co-supervisor, Hein Stigum, for giving me his time in the statistical section of these analyses. You made me understand more how statistical analyses should be done.

I also would like to thank Bernadette Kumar, Haakon Meyer and Folkehelseinstituttet for the HUBRO data. The idea of this research analyses would not have happened without the data. Having the data for “free” is greatly appreciated! Thank you!

Finally, I would like to thank the institute statistician, Lien Diep for her time in helping me do the statistical analyses. She motivated me so much that I have come to understand how to use SPSS. I know I have been giving her a hard time with my complex questions, but our meetings really made me work hard.

Thank you!

Thank you to all my professors and teachers of the International Community Health Institute. Thank you to Vibeke and Line for their untiring support and presence when we needed them.

Thank you to all my friends and my classmates who made my study more exciting and enjoyable. I hope we can continue our good relationship and keep in contact with each other, wherever we are. I will miss all of you!

I also would like to thank Elaine for setting me this deadline because of her first visit to Europe. Her 14th of May arrival in Oslo gave me “pressure” to finish this paper earlier so we can enjoy our vacation. Can´t wait to see you!

Thank you to all my loved ones, my family and Svein who have been patient with me from the start. You all mean everything to me.

Finally, I thank God for the blessing he has given me in my life, especially for the completion of this master thesis.

Oslo, May 2009 Steve Diaz French

(5)

5

List of Tables

Table 1.

Demographic characteristics in Means (SD) and Percentages (N) of the Immigrant groups in the Oslo Health Study 2002. ... 39 Table 2.

Prevalence of Moderate-Severe Musculoskeletal Pain among the Immigrant groups in the Oslo Health Study 2002. [Percent(n)] ... 40 Table 3.

Mean pain scores of the Immigrant groups in the Oslo Health Study 2002. ... 41 Table 4.

Uniivariate association of MSMP to the independent variables. ... 42 Table 5.

Adjusted OR with CI for the association of MSMP and psychological distress of the Oslo Health Study in 2002 by ethnical background. ... 44

List of Figures

Figure 1:

The Dimensions of Pain... ... 9 Figure 2:

Gate Control Theory. ... 11 Figure 3:

The Proportion of men and women with Moderate-Severe

Musculoskeletal Pain, in the Oslo Health Study 2002. ... 45

(6)

6

List of Abbreviations

CAD - Coronary Arterial Disease CI - Confidence Interval

ESEP - Experience of Serious Economic Problem GP(s) - General Practitioner (s)

NPA - No Physical Activity (PA) HSCL-10 - Hopkins Symptom Checklist

IASP - International Association for the Study of Pain

MSP - Musculoskeletal Pain

MSMP - Moderate-Severe Musculoskeletal Pain N - Number (count)

NLP - None or little pain OR(s) - Odds Ratio(s) SD - Standard Deviation

/wk - Per week

(7)

7

Abstract

Objective: Evenwith an increasing immigrant population in Norway, there are still a limited number of studies among the group. Chronic musculoskeletal and psychiatric disorders frequently occur and there is a need to establish the magnitude of prevalence and the strength of association between the two chronic disorders in a local context.

Methods: Cross-sectional data from the Oslo Immigrant Health Study in 2002 were analyzed. Questionnaires were sent to age cohorts, between 20 and 60 years old, among immigrants born in Sri Lanka, Iran, Turkey, Pakistan, and Vietnam.

Results: The results show that neck and shoulders are the most common sites of pain. Women have a higher prevalence of moderate-severe musculoskeletal pain than men do in all five areas of the body. Psychological distress was associated as the strongest predictor of musculoskeletal pain after the adjustment for gender, age, pre-migration factors and others variables in the logistic regression analyses among all five immigrant groups.

Conclusion: Findings from this study support previous studies of the prevalence of musculoskeletal pain and the association between musculoskeletal pain and psychological distress among the minorities in their host country. This also presents the possibility of improving the efforts of the Norwegian health system in providing relevant treatment services for the immigrant population.

(8)

8

1

I

The immigrant population has increased in Norway¹ over the last decade.

Like any other country, Norway has had its fair share of healthcare dilemma related to her increasingly multi-cultural population. For instance, an increasing immigrant population has changed the healthcare landscape of Norway. As an example, there has been an increase in the number of South- Asian population and the increased rate of diabetes mellitus².

However, one knows very little about the health of the new residents of Norway especially with respect to health conditions that may be associated with ethnic origin (1). Successful planning and management of chronic diseases like chronic pain or mental illness requires knowledge regarding the magnitude of these illnesses. This thesis will be the first to discuss the differences in musculoskeletal pain and its association with psychological distress among five immigrant groups (age 20-60) living in Oslo, Norway.

1 http://www.ssb.no/english/subjects/02/03/innvfram_en/

2http://www.fhi.no/eway/default.aspx?pid=238&trg=MainLeft_5895&MainArea_5811=5895:0:15,46 75:1:0:0:::0:0&MainLeft_5895=5825:74058::1:5896:1:::0:0

(9)

9

1.1 Theoretical Background

Pain is a subjective and complex experience that is unique to each individual.

The complexity of a pain experience involves several dimensions (Fig. 1) (2).

According to the International Association for the study of Pain (IASP), pain is defined as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage” (3).

This definition explains that pain is a private experience and therefore complex.

Figure 1: The Dimensions of Pain. (D.J. Magee: Orthopedic Physical Assessment, 4th Edition.

Philadelphia, Saunders, 2002, p.4.) Physiological

Location Onset Duration Etiology Syndrome

Sensory Intensity

Quality Pattern

Affective Mood state

Anxiety Depression

Well-being

Cognitive

Meaning of Pain View of self Coping skills and strategies

Previous treatment Attitudes and beliefs Factors influencing pain

Behavioral

Communication Interpersonal interaction

Physical activity Pain behaviors

Interventions Sleep

Sociocultural-ethnocultural Family and social-life Work and home responsibilities

Recreation and leisure Environmental factors

Attitudes and beliefs Social influences PAIN

(10)

10 Melzack and Wall pioneered the Gate Control Theory (Fig. 2) to emphasize a more incorporated view of the central pain processing at the spinal cord and cerebral levels (4;5). The theory explains the spinal cord as both a passive channel for pain diffusion and as an active modulator of pain signals. The spinal cord can also block nociceptive information to be assessed by the central nervous system (brain). The theory expanded from a purely sensory event to a more complex experience which emphasizes motivational, affective and cognitive aspects of pain experience (3). The sensory, affective, and motivational aspects of pain could activate neural pathways as the effect of pain experience.

Moreover, pain can also be influenced by descending inhibitions from cortical structures. The behavioural-induced reduction of pain is the effect of the descending modulation of the gate that theoretically could block the nociceptive signals at the dorsal horn. On the other hand, pain could potentially increase due to a psychological process, like depression, that can facilitate the ‘opening of the gate’ mechanism at the dorsal horn. Thus, pain experience is influenced by affective (emotional and motivational), subjective (sensory) and evaluative (cognitive) components. Simple reflexes which are usually thought to be spinal in nature are now known to be influenced by cognitive processes (5). For example, when we know the cup we use for our coffee is very expensive, and we pick this cup up while it is very hot, we are unlikely to just drop the cup. Instead, we may jerkily put the cup back on the table, and then tend to our hand by tapping it or blowing it. Furthermore, Melzack developed his theory explaining a pain experience as being unique to every individual (3). This pain experience is influenced by sensory, psychosocial, and genetic factors.

(11)

11 Figure 2: Gate Control Theory: “Conceptual model of sensory, motivational, and central control determinants of pain. The output of T cells of the gate-control system projects to the sensory-discriminative system (via neospinothalamic fibers) and the motivational-affective system (via the paramedical ascending system). The central control trigger is represented by a line running from the large fiber system to central control processes; these in turn project back to the gate-control system, and to the sensory-discriminative and motivational-affective systems. All three systems interact with one another, and project to the motor system.”

(Redrawn from Melzack, R., and P. Wall: The Challenge of Pain. London, Penguin Group, 1988, p.

191.)

central control processes

motivational-affective system (central intensity monitor)

sensory-discriminative system (spatio-temporal analysis)

gate control system

motor mechanism

L input S

T

(12)

12

1.2 Definition of terms

Important terminologies in this study that needs to be defined are:

Musculoskeletal pain

The word musculoskeletal comes from two major anatomical structures in the body, which are “muscles” and “skeleton”. When the origin of pain comes from either the muscles or bone, which may also include soft tissue structures like ligaments and tendons, it is often referred to as musculoskeletal pain (6). Generally we know that pain is an experience (7), musculoskeletal pain in this study was assessed through a self-reported questionnaire. Musculoskeletal pain was divided into none or little pain (NLP) which refers to the group with ‘no pain’ and moderate-severe musculoskeletal pain (MSMP) as the actual group ‘having pain’. Therefore, self-reported pain in this study is considered musculoskeletal in origin.

Psychological distress

The term psychological distress in this study refers to global psychological distress, which was assessed by The Hopkins Symptom check List (HSCL- 10) (8;9). HSCL-10 has five questions each on anxiety and depression (10).

However, it may not necessarily mean that the subjects in this study are depressed or suffering from an anxiety disorder. (More details under independent variables)

(13)

13

1.3 Literature Review

1.3.1 Prevalence and burden of pain

Chronic pain and other health related afflictions represent a ‘black hole’ in the world economy and raise public health concerns to the already burdened health and social care systems (3). The prevalence rates of chronic pain vary widely between 17% to 64% in various population studies (11-15). In Europe, pain prevalence varied markedly between countries – with more than 25% of adults in Norway, Poland and Italy reporting pain, while chronic pain in Spain was only reported by 11% of the adult population (16). However, it may represent more than one third of the population (17;18).

A telephone survey across 16 countries showed nearly 20% of adults suffering from chronic pain (16). The most frequent source of pain was the back (24%) and 35% of the respondents complained that arthritis or osteoarthritis were the most common cause of chronic pain. A Large proportion of reported common diagnoses (i.e. chronic musculoskeletal and arthritic conditions and spine disorders) have been correlated with a high coexisting risk for disability (16;19;20). To illustrate this point, at least 70 million Americans will be diagnosed with arthritis and related disability by 2030 (3).

A similar telephone survey in Sydney, Australia found 22% of the respondents suffer from chronic pain, musculoskeletal complaints were the most common (26%) (16). In both surveys, a high utilization of health services by those affected was noted.

A Swedish study reported that the prevalence of chronic regional pain was almost 24%, and chronic widespread pain was 11.4%. This was found among the general adult population even after controlling for age and gender (21). Two regions in Switzerland were studied for low back pain and it was found that 20.2% of men in the population age 24-34 had had persistent pain for more than seven cumulative days, and 28.5% in the 65-74 years age

(14)

14 group (22). Among women, 31.1% and 38.5% of the population reported pain, in the two groups, respectively.

In Norway, ‘Norgeshelsa’ provide self reported online³ data on somatic pain from ages 16+ (23). The data showed the prevalence of pain among men and women was 19% and 29%, respectively. Hagen et al. found that 17% of the population reported non-inflammatory widespread pain (24). Another study showed, 17.9% reported having pain in five or six areas of the body while 9.8% reported pain in at least seven out of ten possible body areas (25). Rustoen et al. found 58.9% of their participants (ages 19-80) with chronic pain also reported having chronic conditions (i.e. musculoskeletal problems, non-specific chronic pain disorder, osteoarthritis, asthma, gastrointestinal disorders and psychiatric disorders) (26). A follow up study in Nord-Trøndelag found 51% of the respondents had chronic musculoskeletal complaints (i.e. pain and stiffness) (27).

Moreover, pain among children and adolescents has also been found.

In Germany, 83% had experienced pain in the past 3 months and 38.8% of the children and adolescents said that their pain had persisted > 6 months (28). Among those children and adolescents with pain, more than two thirds reported restrictions in daily living activities. In the UK, a survey among pain specialists and general practitioners (GP´s) treating child patients with pain was conducted (29). They found that <5% of children suffered from chronic pain. In spite of this, 22% of the responders (pain specialists and GP´s) reported that the problem has been increasing during the last five years (29).

Other studies suggest, pain increases with age (13;17;18;22;26).

The experience of pain at a young age may indicate the development of chronic or widespread pain later in life.

Lien and his colleagues found musculoskeletal pain among adolescent immigrants in Oslo (30). Girls reported more pain than boys in most parts of the body (head, neck, shoulder, back, and stomach) except for the upper and lower extremities (arms/leg/knee). Even though the differences between

3 http://www.norgeshelsa.no/norgeshelsaen/ (click on ‘key statistics-health status-self reported health)

(15)

15 genders small in the number of pain sites across the immigrant groups, it was nonetheless statistically significant for both boys and girls.

A review of literature amongst the South Asian ethnic minority groups in the UK suggests that South Asian men aged 30-40 reported less musculoskeletal symptoms (14%) than the general population (31%) (7).

However, South Asian women reported more symptoms than the general population (35% and 26%, respectively). The variations may relate to the possible differences in health seeking behaviour and in health conditions.

Alison et al. found that ethnic minorities of South Asian origin (Indians, Pakistanis, and Bangladeshis) in the area of Greater Manchester have a higher occurrence of both regional and widespread pain than the local ethnic population (31). The crude prevalence of musculoskeletal pain (non-specific area) among people age 45-64 was higher than the local ethnic groups.

It ranges from 63% (lowest) among the Indian men and 89% (highest) among the Pakistani women, compared with local ethnic men (53%) and women (55%). There are gaps in the literature regarding the prevalence of musculoskeletal symptoms among the South Asians (or Asians in general) in their host countries (7;31). Thus, one can only make limited comparison using prevalence studies set in the UK and elsewhere.

The development of ‘chronic pain’ has a great impact on related psychological and social functioning (3). Studies in the Netherlands reported chronic pain and other related impairments and disabilities had significant socioeconomic consequences (32;33). These consequences are due to expensive healthcare costs, lost of wages and productivity, and the increasing costs of disability benefits and compensation.

Health related expenditures and lost productivity in the US has been conservatively estimated, at between $70- $120 billion annually (3). About 90 million physician visits per year can be accounted to chronic pain, 14% of all prescriptions, and 50 million lost workdays. Stewart et al. also found that 76.6% of productivity time loss on the job was related to pain and not due to absence from work (34).

(16)

16 Improved health care and advancements in medical technology have lead to an increase in the survival rates for the population. The early diagnosis of illnesses and better medical options available will also have the same effect on persons with life shortening disease. Those diagnosed earlier with terminal conditions may receive the same benefit. The anticipated expansion of the elderly population will further increase the prevalence of pain in the future (3).

1.3.2 Psychological, cognitive and behavioural aspects of pain George Engel proposed the biopsychosocial model in 1977 which incorporates the social, psychologic and behavioral aspects of illness (35).

In one of his articles he discussed various contextual meanings of persistent pain and the importance on how the individual perceives his or her pain (36).

This is in parallel to the conclusion of Melzack and Wall, showing evidence that psychological approaches (i.e. psychotherapy, biofeedback, relaxation, etc.) have powerful effects on pain perception (37), which may change the experience of pain. However, the experience of pain varies from person to person and the individual himself may decide the right approach.

Santos et al. argue that psycho-physiologic pain syndromes and stress- induced pain disorders, as well as physiologic and affective perceptions of pain should be valued as learned reactions under the control of environmental forces (3). Fordyce proposed the ‘operant-conditioning model of chronic pain’, as means to an end in labelling and treating pain behaviours.

Fordyce believed that pain behaviours can be modified by manipulation of rewards and punishments (37); and the way to abolish ‘pain behaviour’ is to stop all rewards (i.e. attention, sympathy from others, medications they want, avoiding chores or people, etc.).

In addition, Turk and Keefe proposed a cognitive and behavioural approach (memory and emotion) (3). They explained that thoughts and beliefs could influence, and be influenced by emotional and physiological responses.

Moreover, the biopsychosocial model has contributed to the development of

(17)

17 a clinically rational school of pain evaluation and management (3). Hence, the experience of pain in this model has incorporated the physical, cognitive, affective, and behavioural components.

Cognitive and behavioural approaches emphasized by Melzack and Wall’s gate control theory, were presented earlier in this paper (4;5). How a person perceives pain, depends on the individual´s beliefs, thoughts, and emotions.

This may influence how pain is experienced. Söderfjell argued that pain is influenced by the individuals’ beliefs caused by reason, duration of pain, precariousness and fear-avoidance (38).

Santos et al. explained the subsequent increased fear and avoidance behaviours in chronic low back pain patients (3). The increased fear levels and disability occurred independently with the pain intensity experienced.

As a result, low back pain patients may be sceptical of moving. Classical conditioning reinforced through operant thinking induces fear of movement.

In order not to extinguish fear, the patient avoids the conditioning anxiety and fear associated with movement. Therefore, decrease in movement and increased expectations of pain, may be due to fear and anxiety.

Pain beliefs, emotions, and passive coping are important affective factors which significantly affect pain response, behaviour, and meaning (3).

Attitudes towards pain are learned early in life, as part of growing up in a particular locality, culture or family (39). These are an important part of any ethnical child-rearing tradition or socialisation. However, this may change over time, as societies go through social and economic development.

Advances may also give rise to new technologies and new methods of pain relief. Therefore, background (i.e. learning, experiences from childhood, culture, and beliefs) may influence the individuals’ understanding of pain entirely or partially.

Korol and Craig suggest that health care providers should be aware of dominant cultural beliefs regarding the medical system and the treatment of illness (40). They quote the South Asian belief about illness as fate, the will of God and ‘karma’. The recovery of illness attributed to the will of Gods and the

(18)

18 skill of the physician. A positive relationship between perceived control over the disease and psychological adjustment was also recognized.

Western health care providers often give false impressions or stereotype, East Asian beliefs (40). East Asian patients tend to somatise distress and are unwilling to report symptoms of psychological conflict. These can be important to the cause and the persistence of their health concerns.

Moreover, the Chinese and East Asians in general, find it more acceptable to receive help for somatic complaints than to receive a psychiatric diagnosis.

Another reason for the non-reporting of psychological symptoms among Chinese patients is the fear of bringing shame upon their families. As a result, others focus on what is interesting to the physicians, which they believe are the physical symptoms that require immediate care.

The association between psychological factors such as depression and anxiety and musculoskeletal pain has been suggested in other studies (3;7;41-44). Recent data have demonstrated that psychological factors predict the later onset of both regional and widespread pain (7). In addition, Macfarlane et al. associated psychological distress with chronic widespread pain (43). Furthermore, Benjamin et al. also found a high prevalence of psychological disorders like anxiety, especially among individuals with widespread musculoskeletal complaints (41).

A review article by Njobvu et al. found consistently, that South Asians and other non-Western cultures tend to express psychological distress through somatisation or somatic metaphors, more often than the native Britons (7).

In addition, cultural variations in the expression of psychological distress make it difficult for primary doctors to recognize the disorders among the South Asian minorities. Furthermore, health questionnaires (i.e. GHQ⁴) used for screening may under-diagnose psychological distress and be culturally insensitive (1). Nevertheless, the association between psychological distress and pain symptoms apply to all cultures (7).

4 The General Health Questionnaire

(19)

19 Merskey argued that anxiety and depression causes muscle tension and eventually pain (45). “We feel one cut from the surgeon’s scalpel more than ten blows of the sword in the heat of the battle” (Merskey: p.625). This quote explains that acute anxiety is liable to increase pain, whereas the high arousal for expected danger may arrest it. He also correlated that over- activity of muscles results in an increased production of waste metabolites.

It may not disappear right away, which can cause pain in the muscles.

He explained that chronic pain patients have reduced levels of maximum voluntary contraction of agonist muscles and an increase co-contraction of the antagonist muscles in the painful area of the body. These are protective physiologically, in order to prevent tissue damage. As a result, the contractions of the high force painful agonist muscles prevented, while favouring the slow movement and reduced speed co-contraction of the antagonists’ muscles. The explanation given earlier regarding ‘fear and avoidance’ (3) model heightens the pain-related anxiety. Avoidance activities serve to promote the ongoing pain, physical de-conditioning and social isolation (3). Nonetheless, pain related to anxieties covers fear reactions across cognitive, behavioural, and physiological dimensions.

Moreover, disturbances in serotonin metabolism is correlated with depression, and on the other hand abnormal serotonin levels have also been associated with pain (43). This may explain the relationship between pain and depression. High levels of pain have been reported by depressed patients (3). They also tend to be less active, report greater disability and life disturbance related to pain, and are more likely to display overt behaviours.

Elevated depression mediates the relationship between higher levels of pain and reduced cognitive functioning. Hence, it is important to understand the complex relationship between depression, chronic pain, and functional impairment.

In conclusion, the number of people suffering from pain and psychological distress is increasing worldwide (46). Chronic musculoskeletal pain has been a major health problem that is common in both developed (47) and developing countries (48). Factors found to be associated with psychological

(20)

20 distress are at work and at home, are low living standards, unemployment, being the victim of a crime or violence, and social factors. The harmful consequences of psychological distress for the individual as well as the community may result in functional disability, economic cost, sick leave and/or work disability (46). Often, under diagnosed by many doctors, psychological disorders are actually common (46;49).

1.3.3 Gender, ethnic and cultural aspects of pain

Various sources (7;11;12;21-24;30;50;51) have demonstrated gender differences in pain among adolescents and adults. Women in general report pain and other symptoms more frequently, not purely those of a particular kind. Berkley suggested the possible explanations to be cultural, biological, earlier knowledge, and psychosocial factors (52).

Women in some cultures welcome pain (for example in childbirth) as a natural experience rather than fearing the experience (39). Many women from Tamil Nadu (India) delivering in the hospitals, have their labour induced and accelerated by drugs such as oxytocin, even if this greatly increase pain during childbirth (53). This is because pain, known as vali – also means

‘strength’ or ‘power’ and is believed to increase the women´s level of sakti⁵ or female regenerative power. Consequently, undergoing a greater pain resulted in greater sakti. In contrast, women in the USA, frequently demanded analgesia during labour (39).

Zola cited American studies among women from lower and upper socioeconomic groups who were asked to report ‘dysfunctions’ in their body (54).

Only a small percentage reported dysmenorrhoea as a dysfunction. Pain like dysmenorrhoea, is not regarded as a dysfunction among many women but rather a natural part of menstruation. Conditions presenting as ‘not normal’

(i.e. pain) that may require medical attention and treatment should be

5 ’The activating principal of life and the principle of endless change that is both celebrated and feared’(Helman, p.173) (53)

(21)

21 defined. It seems, this pain tends to be culturally characterized and may vary over time.

Söderfjell compared gender differences on experimentally induced pain. He found that males are less sensitive, and have higher thresholds, to induced pain compared to females (38). In other cultures or social group, the ability to bear pain among men is more likely expected (39). Thus, Engel points out, pain in this perspective becomes ‘personal’ (55).

Stoicism among the Anglo-Saxon displays ‘stiff upper lip’ in the presence of hardship (39). The ability to tolerate pain without drawing back is a transition from boyhood to manhood as well as a manner of gaining social prestige.

Also in the Great Plains among the Cheyenne Indians, manhood and social prestige is displayed by undergoing a ritual of ‘self torture’ in the Sun Dance ceremony. They suspend themselves from poles by hooks passed through the skin of their chests. Consequently, the acceptance of pain without complaint is an illustration of their strength.

Ethnicity may be a factor to consider in the differentiation of health status (1;56), suggesting that differences identified between individuals in other cultural or ethnic groups are fixed or predisposed (7;39;57). For instance, cultural characteristics such as language and customs are distinct to some groups of people that represent a shared national identity. Therefore, in considering both reporting of pain and pain behaviours, “non-biological”

factors may be of particular importance.

In addition, a review of literature from Giordano, found that socio-medical variations among ethnic groups are dependent on the influence of group and family solidarity (58). Low family dependency during illness means lower ethnic exclusivity, no friendly group solidarity, and no family orientation to tradition and authority. The more cohesive the group, the dependency of the sick individual is also greater. Moreover, social integration within the group is measured by the amount of support sought and secured during illness.

A study was conducted among immigrant women in Canada on how they define their own health (59). It revealed that they are unlikely to talk about the

(22)

22 non-physical aspects of health unless asked about the general contexts of their lives. The family’s health is the final point of settlement for these women. Furthermore, their religious practices and spirituality were essential sources for health.

Zborowski studied reporting behaviours of groups of different ethnic backgrounds (57). He concluded that differences in reporting of pain are culturally bound. These differences are due to beliefs, attitudes and response patterns learned as part of the individuals’ cultural (60;61) or ethnic tradition and/or socialisation .

Moreover, culturally defined languages of distress (i.e. “Aray!” in Filipino) may influence how pain is signalled to others and the type of reaction expected from it (39;60;61). Although a person´s age, gender, and social rank may disagree on this standard, still they are culturally bound (57;62). For example, in societies that value stoicism and resilience, pain is more likely to be expected among men, particularly younger men or warriors (39). As a result, those who fail to follow these norms may meet condemnation or even social sanctions.

Various studies also suggest that individuals from different ethnical backgrounds vary in reporting of pain (7;30;31;39;50;56;57) and co- morbidities such as psychiatric distress (7;30;31;39;41;42;50). These differences may be due to variations in pain thresholds (7;63) and reporting behaviours as pointed out by Zborowski (57).

An interesting review article by Njobvu et al., explained that clinical studies have shown the difference between ethnic groups in terms of pain tolerance and the amount of pain medication required (7). They found that Asian patients have a lower need of post-operative analgesics than the Europeans do. The pain score assessment revealed no difference in both groups (no information given on how they derived the pain scores). Their hypothesis suggests that lower amount of analgesics sedate the Asians easily.

On the contrary, Zatzick and Dimsdale conducted a study (63) on cultural variations in response to painful stimuli. They found that Asians have lower

(23)

23 pain threshold and tolerance than the Europeans. However, they concluded that no consistent evidence could support that ethnic differences existed in the ability to discriminate painful stimuli. They argued that if higher pain tolerance is acknowledged, the occurrence of pain is not entirely due to perceptions or attitudes towards the pain experience. The results of their study suggests that culture profoundly influences pain tolerance, which reflects the behavioural aspects of pain (63).

The impact of pain to the individual and society and its relationship to psychological factors was the motivation for pursuing this study. The aim of this thesis is to contribute additional knowledge and understanding of the association between musculoskeletal pain and psychological distress, especially among the immigrant minorities in Norway. Njobvo et al. stressed in his review paper, that more studies are needed to know the prevalence of pain amongst the ethnic minorities, in order to gain a better understanding about the aetiology of pain related disorders and to provide better health care services (7).

(24)

24

1.4 Objectives

The main objective of this study is to investigate the association between musculoskeletal pain and psychological distress among five immigrant groups in Oslo, Norway.

The following are the specific objectives of this study:

1. To assess the prevalence of musculoskeletal pain among five immigrant groups.

2. To describe the differences in the location of musculoskeletal pain among the immigrant groups across genders.

3. To determine the predictors that might explain the differences in musculoskeletal pain among the immigrant groups.

The following are the suggested research questions:

1. Is there any difference in the prevalence of musculoskeletal pain among the immigrant groups?

2. Is there any difference in musculoskeletal pain complaints between men and women among the immigrant groups?

3. What is the relationship of psychological distress to the prevalence of musculoskeletal pain among the immigrant groups?

(25)

25

2

M

2.1 Study Design

This study will analyze cross-sectional data from the Oslo Immigrant Health Study⁶ (Immigrant HUBRO) of 2002 (64).

2.2 Study area and population

The Norwegian institute of Public Health and the University of Oslo conducted the Oslo Immigrant Health Study in 2002 (65). According to the 2001 population register, 7972 individuals born from 1942-1971 were eligible to participate in the main cohort. Among these, 82 had either died or emigrated prior to the invitation, leaving 7890 for participation. Those reached by mail were 7607. For the young cohort born (1972-1982), 4116

6 Folkehelseinstituttet:

http://www.fhi.no/eway/default.aspx?pid=233&trg=MainLeft_5669&MainLeft_5669=5544:53584::0:

5667:2:::0:0

(26)

26 individuals were eligible to participate in the study. Prior to the invitation, 60 were either dead or emigrated, leaving 4056 eligible individuals for participation. Only 3782 were reached by mail.

2.3 Participation

Main Cohort:

From the 7607 reached by mail, only 3019 gave their written consent and met the inclusion criteria (those who attended the screening or completion of at least one question in either of the questionnaires). Equaling 39.7% in the final response rate reached by mail. The response rates from different countries according to birthplace are 50.9% (Sri Lanka), 32.7% (Turkey), 38.8% (Iran), 31.7% (Pakistan), and 39.5% (Vietnam), respectively.

Young Cohort:

In the young cohort, 3782 were reached by mail, 707 (18.7% of those invited) participated in the study. The participation rates for the 20-30 year olds were 24.7% (Sri Lanka), 18.3% (Turkey), 20.4% (Iran), 15.4% (Pakistan), and 15.2% (Vietnam).

Due to missing data on pain questions, we ended up having 2458 participants in the analyses. Only subjects with complete data on all the variables used were included in the analyses.

2.4 Inclusion criteria

Oslo residents born in Pakistan, Vietnam, Iran, Turkey, and Sri Lanka between 1942 and 1981 were invited to participate in the study. Pakistan has the largest immigrant group therefore only a 30% random sample was invited to participate.

(27)

27

2.5 Exclusion criteria

The Immigrant HUBRO followed the same protocol as the Oslo Health Study 2000-2001 (HUBRO) (66). Individuals who have been previously invited to the earlier study (HUBRO)⁷, belonging to the seven birth cohorts (1940/41, 1954/55, 1960, 1969/70) were excluded.

2.6 Efforts to increase attendance and remaining non

attendees

To increase the participation rates, non-responders from the adult cohort received one reminder between 3-8 months after the first invitation. No reminder was sent to the young cohort. Telephone calls were also made to follow-up the non-responders. In the reminder, mobile screening units were provided in the neighbourhood of those invited visiting 7 sites in the city over the period of 12 weeks (64).

2.7 Data collection method

In 2002, after the approvals and clearances for conducting the survey were given, the local districts and population were informed through mass media with various information techniques about the survey. An invitation was sent to all eligible individuals two weeks prior to the clinical screening; a letter of invitation was sent containing (64):

• Invitation to participate with time and place

• A three-page questionnaire

• Instructions on how to fill out the questionnaire and a letter of consent, to be handed in personally at the screening station

7http://www.fhi.no/eway/default.aspx?pid=238&trg=MainArea_5811&MainArea_5811=5895:0:15,45 62:1:0:0:::0:0

(28)

28

• Information brochure containing the objectives of the survey, content, procedures, etc.

• Map that shows the locations of the screening stations

• All enclosures of this postal package were translated into the five appropriate languages of the target immigrant groups in addition to the official Norwegian version.

Three local districts were selected as screening sites that measured standardised screening procedures like vital signs, blood analyses, height, weight, etc. A supplementary questionnaire was handed out at the survey and could be filled in at the screening site with the assistance of field workers. Field workers that spoke the same language as the respondents were recruited prior to the screening (64). Four weeks after the clinical examination, all participants were informed of the results and received appropriate recommendations according to the HUBRO protocol (67).

2.8 Determining Ethnicity

Only the first generation immigrants belonging to the five selected countries were included in the study. Ethnicity was determined due to country of birth.

The Norwegian population registers determine all residents through a special 11-digit identification code as the basis of their invitation.

2.9 Questionnaire

The main questionnaire for the Oslo Immigrant Health Study was identical to HUBRO (67) that included questions that form part of the larger CONOR⁸ (Cohort Norway) data bank (64) and is available online ⁹.

8http://www.fhi.no/eway/default.aspx?pid=238&trg=MainLeft_5853&MainArea_5811=5853:0:15,28 18:1:0:0:::0:0&MainLeft_5853=5825:56612::1:5857:2:::0:0

9 http://www.fhi.no/dav/C65909BDDD494786B7CFAC0E0E4FF5EE.pdf

(29)

29

2.10 Dependent variable

2.10.1 Musculoskeletal pain

A similar questionnaire in 2000/2001 HUBRO (66) was used to measure musculoskeletal pain complaints. The respondents were asked to report whether they had experienced any pain and/or stiffness in muscles and joints from five different areas (neck/shoulder, arms/hands, upper back, lower back, and hips/legs/feet) in the course of the last 4 weeks. The level of intensity was categorized into “not troubled”, “somewhat troubled”, and “very troubled”.

Musculoskeletal pain data from each region of the body were scored 0, 1 and 2; for no pain (“not troubled”), moderate pain (“somewhat troubled”), and severe pain (“very troubled”), respectively. In obtaining the prevalence of moderate-severe musculoskeletal pain for each of the five areas of the body,

“somewhat troubled and “very troubled” (scores 1 and 2) was merged, while

“not troubled” (score 0) was assigned as none or little pain. The maximum total index score was therefore 10 for the five areas of the body. Mean pain scores were obtained based on the constructed index score (0-10) in this study.

For the logistic regression models, the total index scores were then dichotomized into NLP and MSMP. A cut off score of >2 was labeled as MSMP. The cut off score was used based on the mean pain score of the total sample (gender unaccounted) that is around 3. This provides a division of the group into two (NLP and MSMP) for our main variable musculoskeletal pain (MSP).

(30)

30

2.11 Independent variables

2.11.1 Psychological distress

The Hopkins Symptom Check List-10 (HSCL-10) (8) used to screen for symptoms of depression and anxiety, was developed from HSCL-90 and HSCL-25 (8;9;30;68). Because of its high sensitivity and specificity (8;30;68) , it was used to assess psychological distress in this study.

The HSCL-10 listed various problems asking the responded: (“During the last week (including today) have you felt…”) a) sudden panic for no reason; b) frightened or anxious; c) fainting or dizziness; d) tensed or harassed; e) self- blaming or regret; f) sleeping difficulties or staying asleep; g) sad or dejected;

h) useless or unworthy: i) everything is a burden; j) hopeless for the future.

Each item was rated on a scale of 1 (not troubled) to 4 (much troubled) during the past week. Psychological distress was determined with a score of

>1.85. Strand et al. considered a cut-off 1.85 (for HSCL-10) as a valid indicator to measure mental distress (8;30;69). Subjects that scored above the value of 1.85 in this study were labeled as “psychologically distressed”.

2.11.2 Other independent variables

Age in years was obtained from the year of birth (provided by the population registers) of the participant minus 2002 (year of the survey). This was categorized into four groups (20-30, 31-40, 41-50, and 51-60) and age 20-30 years old was used as a reference group in the univariate logistic regression analyses. However, age was analyzed as continuous variable in the final multivariate logistic regression model.

Gender of the participants was identified through a unique 11-digit identification code from the Norwegian population registers.

(31)

31 For civil status, the participant was asked if they were married, registered partner, unmarried, widow or widower, divorced, separated, separated partner, divorced partner, or a surviving partner. This variable was dichotomized. Married and registered partner were combined together into ‘In relationship’ and other answers as ‘Not in relationship’.

In the question about smoking, respondents were asked if they smoke: (yes, now; yes, earlier; and never). This variable was dichotomized as Yes and No (merging smoking earlier and never), based on the status of the respondent when the data was collected.

The respondent was asked about the physical activity undertaken during spare time in the course of the past year. Physical activities were described as sweating and feel out of breath. The possible answers for the question were (no activity, less than 1hr, 1-2hrs, and 3hrs or more) per week. For the analyses, this variable was not dichotomize.

For pre-migration experience: The participants were asked if they had been injured in war; or tortured (systematic physical or mental maltreatment); both questions were answerable by Yes or No.

The question about experiencing serious economic problems was worded,

‘Have you experienced a serious economic problem?’ the responder could answer Yes of No.

The number of visit to primary doctor, participants was asked how many times (0, 1-3 times, > 4 times) they visited their primary doctor within the past year.

(32)

32

2.12 Statistical analyses

Prevalence between groups was analyzed by a chi-square test, and differences in mean values were analyzed by one-way ANOVA and T-tests.

Logistic regression was used to model effect of the explanatory variables on MSMP.

A Ninety-five percent confidence interval (CI) was set for chi-square tests, one-way ANOVA, T-tests, and odds ratios (ORs) of Musculoskeletal Pain and other covariates included in these analyses.

Data were analyzed using SPSS package 16.0 (70), and P-value less than 0.05 were considered statistically significant.

2.13 Ethical Considerations and Approvals

The Regional Committee for Medical Research Ethics cleared the study protocol that was approved by the Norwegian Data Inspectorate. This study has been completed in accordance with the ethical principles of the World Medical Association Declaration of Helsinki. All the participants of the Oslo Immigrant Study had given their written, signed consent. All the concerned personnel and staff involved in the survey are bound to confidentiality. The study conducted was not anonymous. However, the data has been encrypted to ensure confidentiality for researchers, including the data obtained from registers.

(33)

33

3

R

3.1 Findings

The demographic characteristics of the 2458 participants are summarized in table 1. Mean age was highest among the Pakistani men and women, while lowest among Turkish women and Vietnamese men.

Pakistanis have the highest proportion of men in relationships and this is lowest among the Iranians. The Sri Lankans have the highest proportion of women in relationships and this is lowest among Iranian women. For those in relationship, there is a significant difference between country groups in the proportion of men and women (P< 0.001, P< 0.001, respectively).

The proportion of smokers among men was highest among the Turkish and this is lowest among the Sri Lankans. Among women, smoking was highest among the Iranians and lowest among the Sri Lankans. Smoking was

(34)

34 statistically significantly different among men and among women in the five ethnic groups (P< 0.001, P< 0.001, respectively).

Psychological distress was highest among the Iranian men and women between the five groups and this is lowest among Sri Lankan men and women. Statistical significant difference was found between the males across the five groups and the same result for the females (P= 0.002, P< 0.001, respectively).

The proportions that had experienced torture were highest among the Iranian for men and women, and lowest among the Pakistani men and none among the Sri Lankan women. Torture was statistically significant among men across the country groups and likewise among women, (P< 0.001; P< 0.01, respectively).

For those injured in war, the proportion was highest among the Iranians for men, women, and none among the Pakistani men and women. Significant difference statistically between the country groups was found among men but not among the women (P< 0.01, P= 0.13, respectively) who were injured in war.

Those who experienced serious economic problems, the proportions were highest among the Iranians and lowest among Pakistanis for men and women. The experience of serious economic problems was only statistically significant among men between the five ethnic groups (P< 0.01).

The proportion with no physical activity was highest among the Turkish and lowest among the Vietnamese for men. For women, those with no physical activity were highest among the Pakistani women and this is lowest among the Turkish women. Those with no physical activity were only statistically significant among the women in the five groups (P< 0.01).

The respondents’ who visited their primary doctors > 4 times last year was highest among the Iranian and this is lowest among the Pakistanis for men.

For women, the Turkish have the highest proportion and this is lowest among the Pakistanis. There was no significant difference statistically between men in the five ethnic groups and the same result was found among women.

(35)

35 Comparing the five ethnic groups (not accounting for gender) only no physical activity and those who visited their primary doctor (GP’s) > 4 times have no statistical difference (P= 0.76, P= 0.35, respectively). (Results not shown in table)

Among men across the five groups, the neck/shoulders are the most common site of moderate-severe musculoskeletal pain (table 2). Turkish men have the highest percentage of moderate-severe musculoskeletal pain in all five areas of the body. The Sri Lankans have the lowest proportion of moderate-severe musculoskeletal pain in the neck/shoulder, and lower back areas. Moderate-severe musculoskeletal pain among men in all the areas of the body were statistically different (P< 0.01), except on the arm/hand areas (P= 0.11) across the five groups. Total MSMP was highest among the Turkish and lowest among the Sri Lankans Statistical difference was found between the five groups for the total MSMP (P< 0.01).

For women, the most common site of pain was the neck/shoulders, except for lower back pain, which is the most common among the Sri Lankan women across the five groups (table 2). Sri Lankans have the lowest prevalence of moderate-severe musculoskeletal pain in all five areas of the body. Pakistani women have the highest proportion for those having MSMP in the neck/shoulder area. Moderate-severe musculoskeletal pains in all areas of the body were statistically different across the five countries (P< 0.01). Total MSMP was highest among the Turkish and lowest among the Sri Lankans. The total MSMP was statistically different among the women between the country groups (P< 0.001).

Among the five nationalities, only the Vietnamese have significant difference statistically (P< 0.01) between the proportion of men and women having pain in all the five areas of the body (table 2). Within the Turkish group, only low back pain has no significant difference between genders. For Iran and Pakistan, the proportions with pain in the neck, shoulder, arms, hands, hip, leg, and foot pain areas are statistically different between genders (P< 0.01).

Between the country groups (not accounting for gender), moderate-severe pain in all the areas of the body are statistically different (P< 0.001) (table 2).

(36)

36 For total MSMP, there is a significant difference statistically in the proportion between the five countries (P< 0.001) (table 2; figure 3). There was significant (P< 0.01) statistical gender difference on the total MSMP across the nationalities, except for the Sri Lankans.

Mean pain scores were highest among the Turkish in both genders (table 3).

Between the five ethnic groups, the mean pain score is significantly statistically different (P< 0.001) for gender, age, smoking, and psychological distress, torture, injury in war and ESEP.

Within the country groups, in relationship (civil status) showed a statistically significant difference for Turkish and Pakistanis, (P< 0.01, P= 0.04) (table3).

Smoking showed significant statistical difference for Sri Lankans (P= 0.01) and Iranians (P= 0.02). Psychological distress showed statistical significant difference in all the five groups (P< 0.001).

For the pre-migration factors torture showed a statistically significant difference in the mean pain scores for Iranians, Pakistanis and the Vietnamese (P= 0.03, P= 0.04, P< 0.01), respectively (table 3). Injury in war was significantly different among the Iranians only (P= 0.001).

For those who have experienced serious economic problems, statistically significant differences in the mean pain scores were shown among the four groups (Sri Lanka, P= 0.02; Turkey, P= 0.04, Iran, P< 0.001, and Vietnam, P< 0.01, respectively) except for the Pakistanis (table 3). Physical activity (PA) and the number of visits to a primary doctor showed no statistical difference for all the five groups.

Univariate analyses (logistic regression) were used to show the association of each independent variable with MSMP (>2 of the 0-10 index) as the dependent variable (table 4). This study found several significant (P-value <

0.05) predictors like age, female gender, experienced torture, injury in war, etc., (table 4) to be associated with MSMP.

(37)

37 Without controlling for possible confounding factors, the strongest predictor of reporting MSMP was psychological distress, recording an OR of 7.51 (5.87- 9.61) (table 4). This indicated that respondents who had psychological distress were over 7 times more likely to report MSMP. Being female, the odds of reporting MSMP are 1.65 (1.41-1.94), indicating that females are 1.65 times more likely to report pain than males. The OR for MSMP increases with age. Those belonging to 51-60 age groups have odds of almost three times higher to report MSMP than the 20-30 year old age group.

Turkish immigrants were almost three times more likely to report MSMP than the Sri Lankans (reference group) (table 4); Pakistanis have an OR of 2.05 (1.59-2.64) in developing MSMP than the Sri Lankans, Turkey, 1.8 (2.02- 3.34) and Vietnam 1.5 (1.21-1.92).

Smokers have an OR of 1.56 (1.28-1.90) indicating that they are at least 1.5 times more likely to report MSMP than non-smokers (table 4). Those who have been tortured are 2 times more likely to report MSMP compared to those who have not been tortured. Injury in war predicts the development of MSMP by almost three times over those who have not been injured in war.

Those who have experienced serious economic problems (ESEP) have an OR of 2.5 (1.85-3.31) indicating that they are 2.5 times more likely to report MSMP than those who have not experienced serious economic problems.

Physical activity and visits to a primary doctor were not significantly associated to MSMP. Only variables (table 4) with P-value of ≤ 0.05 were considered significant and were used as the basis of inclusion in the adjusted multivariate analyses model.

The association of psychological distress and MSMP was modeled in a multivariate logistic regression analyses (table 5). Age, gender, and psychological distress were the consistent variables in all five nationalities that showed statistical significance (not shown in the table). Sri Lankans showed the strongest crude OR of 9.18 (5.20-16.19) for psychological distress across the five nationalities. This indicated that Sri Lankans who had psychological distress were over 9 times more likely to report MSMP.

(38)

38 Adjusting for age and gender, only the Sri Lankan group showed an increase in OR to 9.55 (5.39-16.93); and OR was weakest among the Turkish group.

Immigrants from Pakistan showed the strongest psychological distress OR

>18 (4-81) after the adjustment of pre-migratory factors and ESEP. This indicates that those Pakistanis who had experienced torture, injury in war and ESEP were over 18 times more likely to report MSMP, after controlling for pre-migratory factors and ESEP. When other factors were included in the fully adjusted model, ORs increased for Sri Lanka, Turkey and slightly among the Vietnamese. The ORs for Iran and Pakistan decreased and were lowest among the Iranians.

Overall, after controlling for possible confounding factors, the association of MSMP and psychological distress among the five immigrant groups has been shown in the final model (table 5) of these analyses.

(39)

39

3.2 Tables

Table 1. Demographic and background characteristics in Means (SD) and Percentages with Moderate-Severe Musculoskeletal Pain (N) of the Immigrant groups in the Oslo Health Study 2002.

Country of Birth Sri Lanka Turkey Iran Pakistan Vietnam Men

N = 1256 420 191 252 198 195

Age in years 37.2 (7.8) 36.5 (10.3) 37.5 (8.4) 38.9 (11.6) 36.4 (10)

% In relationship† 92.1 (151)* 80.6(83)* 57.1 (68)* 94.6(88)* 66.7(46)*

% Smoking † 24.2 (37)* 59.8 (61)* 50.4 (59)* 41.1 (37)* 43.5 (30)*

% Psych. Distress

(HSCL-10>1.85)† ^25.5⁽³⁹⁾* 39.6 (38)* 47.4 (54)* 26.8 (22)* 31.7 (20)*

% Experienced

Torture† ^16.9⁽¹¹⁾* 5.4 (3)* 58.9 (33)* 2.4 (1)* 28.1 (9)*

% Injured in war† 17.7 (11)* 4.2 (1)* 33.3 (17)* 0* 18.5 (5)*

% ESEPᵇ† 28.1 (18)* 36.7 (22)* 51.7 (30*) 16.3 (7)* 32.3 (10)*

% With No PA҂ per

week 57.1 (28) 68.8 (22) 60.6 (20) 57.7 (15) 47.6 (10)

% Using ͌GP´s > 4

times last yr. 36.4 (12) 25.9 (7) 45.2 (19) 34.3 (12) 42.3 (11) Women

N = 1202 350 201 207 168 276

Age in years 33.9 (8.5) 34.1 (10.1) 35.2 (10) 36.8 (11.2) 34.8 (10.1)

% In relationship† 87.2 (116)* 84.3 (118)* 64.1 (82)* 84.1 (95)* 73.3 (84)*

% Smoking† .8 (1)* 27.6 (37)* 29.8 (68)* 3.9 (4)* 6.0 (9)*

% Psych. Distress

(HSCL-10>1.85)† ^{26.0 (32)}* 48.9 (64)* 51.2 (62)* 40.6 (43)* 41.6 (64)*

% Experienced

Torture† ⁰⁽⁰⁾* 4.8 (3)* 16.2 (12)* 11.6 (5)* 7.2 (5)*

% Injured in war 2.0 (1)* 3.0 (1)* 10.6 (7)* 0* 9.4 (5)*

% ESEPᵇ 25.0 (14)* 23.9 (16)* 33.8 (26)* 12.8 (6)* 23.2 (16)*

% With No PA҂ per

week † ^39.5⁽¹⁵⁾ ^26.5⁽⁹⁾ ^41.9⁽¹³⁾ ^63.0⁽¹⁷⁾ ^47.5⁽³⁰⁾

% Using ͌GP´s > 4

times last yr. 42.5 (17) 47.2 (17) 43.8 (14) 19.0 (4) 26.8 (11) P -values (ANOVA and Chi-square): * Not accounting for gender, statistically significant between the five countries (P < 0.001). †Significant difference (< 0.01) between the ethnic groups (within gender);

ᵇ Experience of Serious economic Problems; ҂ Physical Activity; ͌GP’s-General Practitioners

The association between Musculoskeletal Pain and Psychological Distress among five immigrant groups: Results of a cross-sectional study in Oslo, Norway